Assessing the Sensitivity and Localization Accuracy of Electromagnetic Source Imaging via Computational Simulation...

Area of Science:

• Biomedical Engineering
• Neuroscience
• Epileptology

Background:

• Drug-resistant epilepsy (DRE) in children necessitates neurosurgery for seizure control.
• Accurate delineation of the epileptogenic zone (EZ) is crucial for successful surgical outcomes.
• Noninvasive biomarkers like spikes, detected via high-density EEG (HD-EEG) and magnetoencephalography (MEG), approximate the EZ.

Purpose of the Study:

• To evaluate the sensitivity of HD-EEG and MEG in detecting epileptic spikes.
• To assess the localization accuracy of electric source imaging (ESI), magnetic source imaging (MSI), and combined electromagnetic source imaging (EMSI).
• To investigate the influence of source orientation on detection sensitivity and localization accuracy using realistic simulations.

Main Methods:

• Realistic simulations of dipolar sources with varying orientations were generated from intracranial EEG data of eight children with DRE.
• HD-EEG and MEG data with simulated spikes were generated for each source.
• Sensitivity, signal-to-noise ratio (SNR), and localization accuracy of ESI, MSI, and EMSI were analyzed.

Main Results:

• HD-EEG and MEG sensitivity and SNR varied with source orientation; EEG favored radial sources, MEG favored tangential sources.
• MEG demonstrated higher localization accuracy for tangential sources (3.39±5.21 mm) compared to HD-EEG (9.02±13.09 mm).
• EMSI achieved superior localization accuracy (2.02±3.95 mm) for sources at sulci walls compared to ESI and MSI.

Conclusions:

• HD-EEG and MEG provide complementary information for detecting epileptiform activity, especially when spikes are modality-specific.
• Combined electromagnetic source imaging (EMSI) offers additive value over individual modalities in presurgical evaluation for DRE.
• Findings enhance understanding of modality limitations and capabilities for noninvasive EZ characterization in pediatric DRE.